literature-report

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses a Bash tool restricted to jq, which is used for processing structured literature search results. This is a common and safe practice for data transformation.
  • [DATA_EXFILTRATION]: The skill accesses a local directory .asta/literature/ to read existing research results. There are no patterns indicating that this data is being transmitted to unauthorized external endpoints.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and summarize external academic literature. While the source material (papers) is untrusted, the skill's activities are limited to documentation and structured data processing, with no high-risk capabilities like arbitrary shell execution or system-wide file writes.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 12:06 AM